The video memory for storing image display data is generally stored in a bit map data format. Due to entropy redundancy, the memory space may be wasted. For example, contiguous black or white horizontal lines in the 640×480-dot VGA black-and-white display occupy memory that is equivalent of at least the number of lines times 640 bits.
Not limited to display, image data is generally large. If the image data is treated as is, a large amount of storage area is necessary. Because of the large storage, it is costly. To store the image data, the total volume of the image data is compressed by encoding. Furthermore, the compressed data is processed for efficiency. For example, the compression methods include MH or MMR encoding, JPEG for image and JBIG transformation encoding or arithmetic encoding techniques.
The above described prior art compression techniques are generally complex in encoding process. Since the above encoding techniques require determination and operation processes, the encoding and decoding processes tend to be long. For example, the arithmetic encoding technique such as QM-Coder is complex and slow in processing.
Furthermore, the above encoding techniques have developed in the area of image transmission such as facsimile transmission. Because of the above transmission format, the techniques have a tendency to treat a page as a unit. For this reason, when a portion of the stored image is to be encoded or is to be further edited before storing, it is necessary to encode a large amount of data beyond the portion in the image data and time and efforts have been wasted. The prior art encoding techniques are not generally suited for encoding or editing an arbitrary portion of the image data.
An improved encoding method for redundancy encodes by continuously calculating the run length. Even though a hardware support were available for the task, it would still take several to over ten clock cycles to output a single run length. When the above described prior art encoding techniques are applied to the display data for the video memory, it would take ten or thousand times longer time than the bit-map technique in which the display data is stored and edited in a bit-map format. Such a delay has been a reason for a slow or delayed image rendering process.
For the above reasons, an image processing device is desired to be improved for faster encoding and decoding processes so that the video memory is efficiently utilized.